UMA NOTA SOBRE OS LIMITES TERRITORIAIS DA MATA ATLÂNTICA

Aim: Forest fragmentation is among the principal causes of global biodiversity loss, yet how it affects mutualistic interactions between plants and animals at large spatial scale is poorly understood. In particular, tropical forest regeneration depends on animal‐mediated seed dispersal, but the seed‐dispersing animals face rapid decline due to forest fragmentation and defaunation. Here, we assess how fragmentation influences the pairwise interactions between 407 seed disperser and 1,424 tree species in a highly fragmented biodiversity hotspot.Location: Atlantic Forest, South America.Methods: We predicted interaction networks in 912 sites covering the entire biome by combining verified interaction data with co‐occurrence probabilities obtained from a spatially explicit joint species distribution model. We identified keystone seed dispersers by computing a species‐specific keystone index and by selecting those species belonging to the top 5% quantile.Results: We show that forest fragmentation affects seed dispersal interactions negatively, and the decreased area of functionally connected forest, rather than increased edge effects, is the main driver behind the loss of interactions. Both the seed disperser availability for the local tree communities and in particular the proportion of interactions provided by keystone seed dispersers decline with increasing degree of fragmentation. Importantly, just 21 keystone species provided >40% of all interactions. The numbers of interactions provided by keystone and non‐keystone species, however, were equally negatively affected by fragmentation, suggesting that seed dispersal interactions may not be rewired under strong fragmentation effects.Conclusions: We highlight the importance of understanding the fragmentation‐induced compositional shifts in seed disperser communities as they may lead to lagged and multiplicative effects on tree communities. Our results illustrate the utility of model‐based prediction of interaction networks as well as model‐based identification of keystone species as a tool for prioritizing conservation efforts. Similar modelling approaches could be applied to other threatened ecosystems and interaction types globally.

Habitat fragmentation reduces biodiversity and affects ecological processes that are fundamental for maintaining ecosystem services. We investigated how landscape structure—percent forest cover, patch density, percent cover by edge, perimeter-area ratio, and spatial heterogeneity—affects the diversity of mammalian carnivores at multiple extents within 22 Atlantic Forest landscapes. We hypothesized that 1) species richness of carnivores is positively related to forest cover; and 2) the occurrence of species will vary according to its sensitivity to forest loss and its preference for forest or open areas. Species richness, composition, and occurrence of mammalian carnivores were correlated with several landscape structure metrics. Due to a high correlation among the metrics, we adopted forest cover as the principal predictor variable. We compared a forest cover model to a null model using Akaike Information Criterion (AICc), and evaluated other fragmentation metrics using a redundancy analysis. Carnivore species richness was positively related to forest cover and negatively associated with other fragmentation metrics. However, the responses to landscape structure differed among species, possibly due to their differences in habitat use. Landscape configuration is an important factor influencing carnivore species composition. Forest cover can explain some, but not all, carnivore species occurrence. Our results reinforce the protection of forests as fundamental to the conservation of carnivore species and the ecological processes in which they participate. O processo de fragmentação de habitat reduz a biodiversidade e afeta processos ecológicos fundamentais para a manutenção de serviços ecossistêmicos. Nós investigamos como as métricas de estrutura da paisagem—cobertura florestal, densidade de manchas, cobertura porcentual de borda, relação perímetro-área e heterogeneidade espacial—afetam a diversidade de mamíferos carnívoros em múltiplas extensões em 22 paisagens da Mata Atlântica. Orientamo-nos pelas seguintes hipóteses: a) a riqueza de carnívoros é relacionada positivamente com a cobertura florestal; b) a ocorrência das espécies varia de acordo com a sua sensibilidade à perda de floresta e à preferência por florestas ou áreas abertas. Riqueza, composição e ocorrência de mamíferos carnívoros foram relacionadas com várias métricas da paisagem. Devido à alta correlação entre as métricas, adotamos a quantidade de floresta como principal variável preditora. Nós comparamos modelos de cobertura florestal com modelos nulos usando o Critério de Informação Akaike corrigido (AICc) e avaliamos outras métricas da estrutura da paisagem usando análise de redundância. A riqueza de espécies de carnívoros está positivamente relacionada à cobertura florestal e negativamente associada à fragmentação. Porém, as respostas à estrutura da paisagem diferiram entre as espécies, possivelmente devido às diferenças quanto ao uso de habitat. Configuração da paisagem é importante para explicar a composição de espécies de carnívoros. A cobertura florestal pode explicar a ocorrência de algumas espécies, mas não todas. Nossos resultados reforçam que a proteção das florestas é fundamental para a conservação das espécies de carnívoros, e os processos ecológicos nos quais participam.

Aim: To assess the distributional patterns of fruit‐feeding butterfly species richness in Atlantic Forest (AF) based on stack species distribution models (SSDM); to evaluate the relative contribution of climate and landscape in the patterns of butterfly species richness; and to recommend conservation guidelines for AF regions based on the obtained results.Location: Brazilian Atlantic Forest.Methods: We used SSDMs to generate potential distribution maps of butterfly species richness in AF. Our dataset comprises 7,062 records of 279 species, distributed across 122 local communities. We built richness maps based on climate and landscape models and analysed them separately to understand their contribution to the pattern of richness distribution. Then, we combined both models to build a consensus map. Finally, we designed specific conservation strategies based on richness categories distributed on the consensus map.Results: The regions with high‐predicted species richness for both climate and landscape models supported the main endemism locations described for other AF taxa in previous studies. Both models predicted intermediate values of species richness for most regions, but the patterns of richness distribution were not fully congruent. High species richness based on landscape metrics was frequently concentrated in the south. Furthermore, locations suitable for high species richness according to climate are near to the endemism centres, but also to dense urban centres, highlighting the potential impacts of the landscape modification on butterfly species richness.Main conclusions: The distribution of fruit‐feeding butterfly species richness and endemism reflect the classic division of AF in different sub‐regions according to the endemism centres proposed by previous authors for other taxa. The availability of forested habitats seems to be a determining factor for the maintenance of high species richness in AF. Therefore, the loss of natural forest remnants due to landscape modification is the primary threat to butterfly diversity.